Spectrum, Intensity, Health and Aesthetics

[Crit21]

[Crit21]

[Crit21]

[postaldave]

subscribed....

noticed you said red was responsible for bleaching but your data shows it's required.
how would your measure how much need in the tank?

loving the level of research you are doing. been wanting to switch to LED but suspected that these issues had to be addressed first.

[nemosworld]

great thread!

[pfi$h]

Very well done! Thanks for posting this.

[Crit21]

Originally Posted by postaldave

noticed you said red was responsible for bleaching but your data shows it's required.
how would your measure how much need in the tank?

Good question. Let me clarify what the data is saying.

First, red light isn't necessary for photosynthesis in corals. At best, it's a minor contributor to photosynthesis in corals living in the shallowest water. You don't need full spectrum light to grow corals. Keep in mind that corals below 10 meters depth receive virtually no red light. They grow and thrive in the blue region.

There's an ongoing "discussion" between Dana Riddle and Sanjay Joshi regarding red light, coral bleaching and the effect of red light on coral photosynthesis. Dana has done extensive research in this area, while Sanjay has done virtually none. Dana's opinion is that supplementing with red light is somewhere between "inadvisable" to "damaging".

I asked Dana about the red spike in the photosynthetic response graph. His response: "No one disputes zooxanthellae absorb red light. It is the amount absorbed over an extended photoperiod within an aquarium [that is the issue]." His next round of experiments here will look at the amount of red light in a shallow tide pool.

As far as supplementing with red light, he had this to say: "I've bleached many corals and zoanthids with red light, including Pocillopora meandrina, Porites lobata, Pavona varians, and an unidentified zoanthid. P. meandrina and Porites lobata were collected in shallow water. Pavona varians was collected in shallow water, but lives in crevices and overhangs to avoid light (it is in essence a deep water coral). The zoanthids were collected in a surge area and are exposed to the atmosphere for a good portion of the water."

I've pointed to several studies that relate red light to negative effects on coral health and growth. If you insist on making sure your corals are receiving red light, consider that white LEDs supply red wavelengths.

[Crit21]

If you're counting on the reef lighting industry to conduct research into the spectral needs of corals, don't hold your breath. They don't have the time to conduct studies on coral growth, pigmentation, health while their competitors are pumping out the latest LED fixtures that promote fluorescence and make clown fish "pop". IMO, they're advancing LED design in every area but the one that matters.

When you look at some of the LED manufacturers' claims regarding chlorophyll b, you'll see what I mean. They're throwing out scientific jargon in an attempt to lure buyers, when in fact, they're exposing their lack of knowledge. Corals use chlorophylls a and c; plants use a and b.

[Crit21]

This is one of my favorite studies, since it examines coral growth under red, green, blue and clear filters of varying levels of light filtration in the coral's natural water environment. http://jeb.biologists.org/content/212/5/662.full.pdf

[michael_in_nc]

Have you built a new fixture yet, or still doing research?

I think you have some good points on the color spectrum. I have modified my plan based on your work and my initial test build. I had a few questions / thoughts:

1. I like the shimmer effect I get with my halides, but my test fixture had significant color banding. Maybe the color banding wouldn't be as significant with a different LED color mix?
- My plan was to use a diffuser for the color LED's and 60 deg optics on the white LEDs. (someone suggested that) If I remember, you were not using a diffuser, how do you deal with the color banding?

2. What color spectrum devices are you planning on using? I was looking at Violet (410-420nm), RB (450-465nm), Blue (465-485nm), Cool White, Cyan (phillips rebel; 490-520nm). I would probably use a fairly even mix, maybe less of the cyan with Arduino programmable control over the individual color LED's drive level.

3. I'm looking at building "pods" of lights. I would use 9 to cover my 250G tank (84x24 footprint).

I'm working on the electronics for the drivers and control and will order LED's and heatsinks once I have that ready. I could always start without the diffusers and add them in (or not) once it's running, but will probably plan for it up front.

Any inputs?

[Crit21]

I don't need to do any more research for my LED design, but as I read more data I'll share what I find. I haven't built it yet, but building an LED fixture is easy.

As far as banding, it's best if you can avoid bands of LEDs of the same color. Distribute the LEDs in a mixed pattern, like a zig zag, but keep each color on the same circuit so you can control each color independently. LEDs have great shimmer. Each one is a point light source. Diffusers will eliminate shimmer.

You choose your lens angle based on two things: 1) The distance between the LED and the upper corals, and 2) the distance between each of the LEDs. It's easiest if you draw it out to scale on paper, then play with the LED spacing and cone angles (use a protractor) until you get the effect you want. You want the beams to overlap so you don't get color banding or spotlighting.

I'll be using the LEDs shown on one of the graphs near the end of my first set of slides. Where did you find your violet?

You'll have a couple of problems to overcome with pods of lights. Closely spaced LEDs will give the appearance of spotlights. The second issue will be cooling. How will you address them?

[michael_in_nc]

Rapid has violet LEDs. I was planning to buy from there.

The pods will be pretty large so the spacing shouldn't be too bad. I have scale diagrams of my plan. The pods will be mounted high 12-15" off the water surface which should reduce spotlighting. I know someone with a sandblaster and they said it would be possible to mask off regions of the diffuser and also do some test runs with very light etching so there is much less diffusion.

It will be about 20 LEDs per heat sink. Most won't be driven near full power, so maybe 40-50W total per heat sink. I was going to put fans on the heat sinks and temp monitors on the controller.

[Crit21]

I still don't understand the need for a diffuser at all. I had nothing but lenses over mine.

[michael_in_nc]

What mix of colors did you have? RB/Blue/CW? I think the issue is the wider range of light spectrum we are discussing. Now, my test fixture did have green & red, but I still think it will be an issue at some level with more wider spectral coverage with such narrow spectral output bandwidth of LEDs.

I don't think spacing alone is the key as I had all of my LEDs in a very small footprint on the order of the typical spacing used for just single LEDs. I was using 60deg optics on all LEDs. The tighter spacing should put the light outputs well overlapped, so they should "blend". I think the issue is the light has slightly different angles based on the separation, so anything that makes a shadow could block one color and not the others or vice versa. The result is a spot of color that kind of dances around with the agitation of the water surface.

I have built it (on a small test structure) and observed it, so that's why I'm so concerned about addressing it before I build the full fixture which will be expensive. Now, I hope you're right and I don't need it, but I'm just planning for the worst. I think your suggestion of dropping the red and pulling the green in to cyan will help considerably because the LED spectral bandwidth will be much tighter and have more overlap between individual colors.

Appreciate the inputs...it will make my end result better. Will definitely drop the red, pull in the green to cyan, and adjust my balance of LED colors. Hopefully that will help my diffuser concerns as well...

Another question, do you use all cool white or use any other color temps on those or a mix? I'm thinking it won't make much of a difference with the other color spectra LEDs....

[bambam918]

really enjoyed the read, thanks for the post.

[Crit21]

Originally Posted by bambam918

really enjoyed the read, thanks for the post.

Thanks. I'm just trying to get the ball rolling on the discussion. Everyone is invited to contribute.

[Crit21]

Michael_in_nc, My last build was cool white, RB and blue in equal numbers. I had them spaced about 3" on center in both directions. I was using 60 degree lenses. I had very few shadows.

My build this summer will include all of the colors I included in the slide in post #18. You'll notice that there's plenty of overlap. Monochromatic LEDs don't have as narrow of a bandwidth as you think. Notice the curves for each color.

[Crit21]

Yes, I used all cool white Crees.

Going with cyan will help with color depth, but aesthetics is a matter of personal taste. Some prefer more or less of each color, which is why each color should be on a separate dimmable circuit. I've heard from a couple of LED users that green really makes things look funky. Green also does nothing for photosynthesis, because it's not absorbed. That's why leaves are green--chlorophyll reflects green wavelengths.

[Crit21]

I'm probably going to use more LEDs of each color because I may not be turning each color to 100%. I haven't tried this color combo before, so I want to play with the intensity of each for aesthetics. I know I'll have a ton of PAR/PUR with these colors, but they may not look good at full intensity. If I turn them down, I still want to make sure the corals are getting enough PAR.

[Crit21]

I should make one point. I'm not recommending that anyone go out and build an LED system tht covers the entire 400-500nm range. It's probably not any better for photosynthesis, because the antenna pigments can collect energy from a wide range of wavelengths, although some wavelengths produce a much greater photosynthetic response than others. As you may have experienced, you can get some pretty significant growth from just blue and RB LEDs. I can predict two significant advantages from using a broader waveband of light. First, you'll cause more pigments to fluoresce. Second, you'll get a different aesthetic look with more wavelengths of light, and will have an opportunity to adjust those different colors to get the look that you prefer.

[michael_in_nc]

What are people typically using in terms of number of LED's for the typical RB/B/CW setups? Number per square inch or something like that? Just looking for a general rule of thumb to double check that I'm in the right ball park.

I would generally plan to keep those numbers the same and just augment with the other additional colors, keeping the ability to control those independently as Crit mentions.

[Jpellica]

tagging on this one Thanks

[OneReef]

I tried to read all the slides and keep up, but got lost along the way, lol. Good info though, good research. I guess what I am trying to extract from your data is in regards to my AI Sol units. They have white, blue and royal blue. Each color independently controlled. According to your info I would get better coloration and growth by using more intensity in the blue and royal blue LEDs and less with the whites? One of the problems I think a lot of us have with LEDs with all the adjustability is what setting works best for our coral color and growth. I have mostly SPS. What's difficult is knowing if you're using too much or not enough light, and what is the best spectrum to dial in.... With my PAR meter, Ive found that my SPS near the top get around 350-400 and any more than that they start to pale. So your research makes me wonder If I should dial back the white a bit and use more blue. I currently run them at a 1:1:1 ratio. For example, 70%W, 70%B , 70%RB. Would I be better off going 50/75/75? That is the heart of my questioning.

[drtrash]

I added a supplemental colored LED to my sol blues and it has been nothing but positive. Been running for a months and colors look better and are responding better to light spectrum. I added 10 violet 2 green & 2 red, blended with a few royal blue to tone it down. I used a diffuser screen to final blend color. Visually tank was missing definition of color, so very happy with results. I would of liked to test each color seperately to determine which wavelength had the greatest impact but it was a package deal no time for exp. Wife thinks I crazy already. Red is a strong color but does have some pop quality that is hard to deny, in moderation. I know its aesthetics but thats why I have it, also cheaper than diving

[Crit21]

Originally Posted by michael_in_nc

What are people typically using in terms of number of LED's for the typical RB/B/CW setups? Number per square inch or something like that? Just looking for a general rule of thumb to double check that I'm in the right ball park.

I would generally plan to keep those numbers the same and just augment with the other additional colors, keeping the ability to control those independently as Crit mentions.

There's recent data out there by Sanjay where he measures the LED watts per square inch (http://www.advancedaquarist.com/2012/1/aafeature). I'm aiming for .20W/sq in, which should give me enough extra total PAR that I can either run all LEDs below 100%, or turn down one or more colors to get the aesthetic look I want without under-exposing the corals.

There's more to it than that though. The LEDs won't be mounted all the way to the edges of the tank; they'll be pulled back 6-9" so the light doesn't bounce off the sides of the tank too much. I'll post a diagram from my last DIY LED project to clarify what I'm saying later tonight if I can find it.

As far as numbers of each LED, I think I used about 25% blue, 25% royal blue, 50% white. I ended up dimming the whites because they were so overpowering. I'm thinking that if I were to do it again I'd use 1/3 of each.